Open AmandaFranklinRyan opened 10 months ago
Maybe we could have 3 versions of the data and see what works best with the models:
A very interesting plot! I think we should add this one to our final report. And it sounds like a reasonable idea to create 3 versions of data and based on this decide which data set we include.
I'm not sure how we should deal with the correlated features in the data set. Clearly micro rating is correlated with the different individual micro ratings and the variables themselves, but how do we know which ones to keep?